RecipeMatching reference (Qwen3-Embedding-0.6B)

A checkpoint from the paper Recipe-Matching, Not Equivalence (Ali Habibullah, Mohammad Alshiekh, Yazan Alshoibi, Salman Khan and Naeemullah Khan, 2026), shared so the paper's evaluations of this model can be re-run without retraining. It is not one of the three models the paper releases as its main artefacts. Code, data and results: https://github.com/KAUST-Academy/recipe-matching-not-equivalence.

What it is. The paper's reference model: Qwen3-Embedding-0.6B fine-tuned (all weights) on 6,145 rows drawn from data/llm_pairs_unrelated_casnegs/pairs.jsonl, which pairs D4's LLM restatements, written under a prompt unrelated to MathNet-Retrieve's, with the verified arm's counterexampled negatives for the same source problem. No benchmark prompt appears anywhere in its training data. Seed 42; the full training record, command line included, is run_config.json in this repository, and every training file named here is in the GitHub repository.

Results of this checkpoint (seed 42, R@1 / R@5 / R@10; the paper's main tables report eight-seed means (seeds 42–49)):

Evaluation	R@1	R@5	R@10
MathNet-Retrieve easy tier (15,000 queries, 117,088 documents)	46.51	88.06	93.75
MathNet-Retrieve medium tier	4.34	42.20	60.62
MathNet-Retrieve hard tier	10.81	60.27	75.58
Cross-language duplicates (strict)	77.61	82.95	84.73

Usage. Encode queries with the model's query prompt, the setting every number above uses.

from sentence_transformers import SentenceTransformer

model = SentenceTransformer("KAUSTAcademy/RecipeMatching_Reference_Qwen3-Embedding-0.6B")
q = model.encode(["Find all real x with x^2 = 2x."], prompt_name="query")
d = model.encode(["Determine every real solution of x^2 - 2x = 0."])
print(model.similarity(q, d))

Citation

@misc{habibullah2026recipematchingequivalence,
      title={Recipe-Matching, Not Equivalence}, 
      author={Ali Habibullah and Mohammad Alshiekh and Yazan Alshoibi and Salman Khan and Naeemullah Khan},
      year={2026},
      eprint={2609.31927},
      archivePrefix={arXiv},
      primaryClass={cs.IR},
      url={https://arxiv.org/abs/2609.31927}, 
}